Both children and adults can have cardiac problems. Congenital problems predominate in childhood, whereas adults are more likely to suffer from conditions associated with age. In both cases, however, accurate diagnosis depends on access to health care and the availability of trained specialists. In two recent studies, Arnaout et al. and Yao et al. showed how machine learning can supplement specialist care in both pediatric and adult cardiology settings. Arnaout et al. analyzed fetal ultrasound images to detect congenital heart disease. Yao et al. used machine learning in conjunction with electrocardiogram imaging to detect adults with low ejection fraction (a measure of the amount of blood that the heart succeeds in pumping), which is a risk factor for subsequent heart failure. In each case, the technology should help to improve diagnostic accuracy and access to appropriate treatment.
Nat. Med. 27, 815, 882 (2021).